Towards Unsupervised Crowd Counting via Regression-Detection Bi-knowledge Transfer
This addresses the problem of counting people in crowds without labeled data for researchers in computer vision, though it is incremental as it builds on existing transfer learning and knowledge distillation techniques.
The paper tackles unsupervised crowd counting by transferring knowledge from regression- and detection-based models in a labeled source set to an unlabeled target set, using mutual transformers and iterative self-supervised learning, achieving substantial improvements over state-of-the-art methods on benchmarks like ShanghaiTech, UCF_CC_50, and UCF_QNRF.
Unsupervised crowd counting is a challenging yet not largely explored task. In this paper, we explore it in a transfer learning setting where we learn to detect and count persons in an unlabeled target set by transferring bi-knowledge learnt from regression- and detection-based models in a labeled source set. The dual source knowledge of the two models is heterogeneous and complementary as they capture different modalities of the crowd distribution. We formulate the mutual transformations between the outputs of regression- and detection-based models as two scene-agnostic transformers which enable knowledge distillation between the two models. Given the regression- and detection-based models and their mutual transformers learnt in the source, we introduce an iterative self-supervised learning scheme with regression-detection bi-knowledge transfer in the target. Extensive experiments on standard crowd counting benchmarks, ShanghaiTech, UCF\_CC\_50, and UCF\_QNRF demonstrate a substantial improvement of our method over other state-of-the-arts in the transfer learning setting.